J . Org. Chem., Vol. 44, No. 13, 1979 2203
Activation of Reducing Agents
Activation of Reducing Agents. Sodium Hydride Containing Complex Reducing Agents. 11.' Regioselective Reduction of a,p- Unsatura ted Ketones L. Mordenti, J. J. Brunet, and P. Caubere* Laboratoire de Chtmie Organique I , ERA C N R S No. 476, L'nicersitd d e Nancy I , Case Officielle 140, 54037 N a n c j Cddex, France Received November 21. 1978
Complex reducing agents NaH-RONa-MX, (MCRA) and NaH-RONa-MX,-MgBrn (MCRA-MgBrn) are described as very efficient reagents for the 10096 regioselective reduction of a,$-unsaturated ketones in THF. Selective ;.,4 reductions were observed with MX, = Ni(OAc)z while selective 1,2 reductions were obtained with MX,, = ZnCI?. Finally. control experiments pointed out some unexpected reducing properties of NaH-ZnCln-MgBrs systems. T h e selective reduction of either the double bond or carbony1 group of t\ ,&unsaturated ketones has been investigated by other workers'-l without complete success. Our work on the reduction of saturated ketoness by complex reducing agents (MCRA), NaH-RONa-MXn6,7 suggested trial of these cheap, safe, and easily handled reagents for the regioselective reduction of enones.
11).
R e s u l t s and Discussion In previous work, we have shown that NiCRA (NaHRONa-Ni(0Ac) 2 ) was able to reduce carbon-carbon double bonds8 and t h a t NiCRA-M'Xf, ( M I X k = alkaline or alkaline earth metal salt) were good reagents for the reduction of saturated ketones.;' On the other hand, it was found that ZnCRA (NaH-RONa-ZnC12) exhibited very poor reducing properties toward carbon-carbon double bondsg but led to good results for the reduction of carbonyl groups, specially when mixed with alkaline or alkaline earth metal salts5 From these observations, one might conjecture that nickel-based reducing agents could be good candidates for the conjugate reduction of enones and zinc-based ones for the regioselective reduction of the carhonvl group of enones.
" 1, R ' =. Me; R 2= H ; R 3 = H;R4 = Me 2, R ' = Me; R 2= Me; R 3 = H ; R4 = R.le 3, R ' = Me; R * = Me; R' = Me; R 4 = Me
4, R ' 5, R' 6 ,R '
= = =
t-Bu; R 2 = H;R' = H; R4 = f-Bu C,H,; R 2 = H; R' = H; R4 = Me C,H,; R: = H ; R 3 = H; R4 = C,H,
0 7, R' = H; R2= H; R3= H 8, R ' = Me; R 2 = hle; R 3 = Me
carbonyl group; (ii) addition of alkaline or alkaline earth metal salts generally improves the CRA's performances; (iii) 40 "C is an adequate reaction temperature, allowing fast reductions; however, with the most sensitive cy, P-unsaturated ketones, reduction must be carried out a t 20 "C to avoid too large formation of byproducts. The optimum conditions thus defined were applied to the reduction of enones 1-9 in T H F (Table -.
0 9
Enones 1-9 were chosen for this study. In our opinion, they are representative of the different situations which may be encountered. First, a short systematic study was performed on three representative ketones 2 , 4 , and 8. From the results (summarized in Table I), the following observations were made: (i) as expected from our hypothesis, NiCRA led essentially to conjugate reduction while ZnCRA led to selective reduction of the
0022-3263/79/1944-2203$01.00/0
Variation of the nature of the metal induces a dramatic variation of the reducing properties of CRA. Although reduction yields were not quantitative in all cases, it is noteworthy that the regioselectivity was always loo%, with both cyclic and acyclic enones. Results of Table I1 deserve some further comments. First, in a few cases, reductions were accompanied by formation of intractable byproducts. Second, in most cases, reduction by
0
II
C,H CHCHICC,H
I
C,H,CH,CHCC,Hj
II
0 10 nickel-based reagents gave small amounts of saturated alcohols if the reaction times were increased. Finally, the reduction of 6 was accompanied by the formation of diketone 10 (50% yield) arising from Michael addition of the saturated ketone enolate to 6. Comparison of these results with those described for usual reagents showed that CRA are very efficient reagents for the regioselective reduction of a,P-unsaturated ketones, leading either to saturated ketones (with nickel-based reagents) or to allylic alcohols (with zinc-based reagents). Control experiments on the reducing ability of CRA's components were performed with ketones 2, 4, and 8 which were easily reduced by either NiCRA or ZnCRA. It was found that NaH, NaH-t- AmONa, NaH-Ni(OAc)p, NaH-MgBr?, and NaH-t -AmONa-MgBrz systems exhibited very poor reducing ability toward enones 2,4, and 8. Except for 2 (which was destroyed when treated by NaH, NaH-t- AmONa, NaH-Ni(OAc)z, or NaH-t- AmONa-MgBra), the ketones were quantitatively recovered in all other cases. I t was also found that NaH-Ni( OAc)a-MgBrz effected the conjugate reduction of 8 in 20 h (instead of 30 mn with NiCRA-MgBrp) while 2 and 4 were quantitatively recovered after 20 h a t 20 "C. Although very poor reducing properties were found for NaH-ZnC12 (see Table 111) and NaH-MgBrz (vide supra), their mixture (NaH/ZnClz/MgBrz 40/10/10 mM) was able to reduce enones to a#-unsaturated alcohols in very good yields. However, it must be noted that reaction times were twice those (C 1979 American Chemical Society
2204 J . Org. Chem.. Vol. 44, No. 13, 1979
Mordenti, Brunet, a n d Caubere
Table I. Reduction of Enones with MCRA and MCRA-MgBr:, in THF
B
A ketone
reducing system
T , "C
2
NiCRA" NiCRA-MgBr2d ZnCRA? ZnCRA-MgBrzd NiCRA NiCRA-MgBrs ZnCRA ZnCRA-MgBrz NiCRA NiCRA-MgBry ZnCRA ZnCRA-MgBrl
20 20 20 20 20 20 40
4
8
reaction time
('0
2h 15 min 2h 2h 2h 2.5 h 2h 3h 4.5 h 30 min 1 hh 3.5 h
40 40 40 40 40
C
Ah
' (1
Bh
0.0
ch
98 95 10Ll;ii 98 95 95 95 9$5 91 93
1
5
68A 90
traces
40-45 ml. Absolute yields determined by GLC analysis with internal standards. NaHlt-AmONalNi(OAc)./ketone= 40/10/10/10 mM. MgBr2, 2 T H F (10 mM) (see ref 5 ) was used. NaHlt-AmONalZnClelketone = 4O/lO/lO/lO/ mM. No recovered starting enone. I: 30% unreacted enone was still present. No further evolution (20 h).
Table 11. Reduction of a,@-UnsaturatedKetones by NaH-t-AmONa-MX,-MIX1,
\
/
11) NaH-tAmOSa-MXna-M
C-
XI, THFb
1?)HO
, ' /CH-cH \ c-
II 0
+ \ c=c / /
\
in THF
+ CH-
0
I
OH
OH B -
1
2 3 4 5
6
7 8 9
____MXn Ni(OAc):! ZnCI:! Ni(OAc)2 ZnC12 Ni(OAc)2 ZnC12 Ni(OAc)n ZnCI:! Ni(OAc):, ZnClp Ni(OAc)? ZnCl:! Ni(0Ac)Z ZnC12 Ni(0Ac)Z ZnC12 Ni(0Ac)n ZnCI:,
MIX',
MgBrZd MgBrZd MgBr2d MgBrzd MgBrZd MgBrZd MgBrZd MgBrZd LiI' MgBr2d MgBrzd MgBr:!d LiI' MgBr2d MgBred MgBr2d MgBrZd MgBrzd
T , "C
reaction time
20 20 20 20 20 40 20 40 20 20 20 20 20 20 40 40 20 40
30 min 30 min 15 min 2h 4 5 min 2h 2.5 h 3h 15 min 20 h 3.5 h 4h 30 min 6h 30 min t3.5 h 1.5 h 3h
CH-
I
II
A ketone
/ \
'CH-CH /
wo
A
(I),,
n
c 0.0
c
80 ;it5-60
98 98
traces traces traces
99 (65) 93 (7,5 1 93 (91) 93 (901 85 (83) 4
